Young Earth Creationist dynamics.

[Valkhorn]

Facepalm!

Only a creationist would balk at a 3-5% uncertainty or margin of error all the while accepting an age of the Earth that has a margin of error probably well over 100,000,000% to the actual value.

 

[RickG]

Hey Rick, this isn't my field so I only jumped in when I saw Z getting statistics wrong (again), but I do have a question, what is the measure of error for this field? Are they talking standard deviation, standard error, 95%CI, or something else? Also in biology we give the absolute values rather than relative.

I can understand how a layman could get confused the way they jump between accuracy, error and uncertainty.

You did a great job explaining to Zaius what those percentages actually meant.

It is all dependent upon the application. Statistics or more specifically, statistical analysis, is one of those fields that is grossly misrepresented and easily done so because few people have any idea of the numerous statistical methods and their applications beyond averaging numbers or determining a percentage.

In the paper of question they are using algorithms but don't discuss the exact details. That's one of the reason that I mentioned it was layman friendly. Anyone wanting to know exactly the methods and even raw data can obtain that information by contacting the authors. This is standard practice, especially withing the climate science community. It's actually a secondary level of peer review for other scientists who either want to verify the results or question them.

Anyway to get back to your question, many percentages of uncertainty or computation of error bars are derived from various regression analysis or methods using least squares for multiple variables. Again, the bottom line is what is appropriate for a specific application. There is no one glove fits all. One of the best sites I've come across for statistical methods and explanations of their uses is Credit Scoring, Data Mining, Predictive Analytics, Statistics, StatSoft Electronic Textbook.

 

[Zaius137]

Hey Rick, this isn't my field so I only jumped in when I saw Z getting statistics wrong (again), but I do have a question, what is the measure of error for this field? Are they talking standard deviation, standard error, 95%CI, or something else? Also in biology we give the absolute values rather than relative.

I can understand how a layman could get confused the way they jump between accuracy, error and uncertainty.

Maybe you participants should examine the statistical terminology before patting yourselves on the back…

"accurate to approximately 3%"

that is covered here… "Accurate to" and "confidence levels" are distinct terminologies.

Statistics Glossary - confidence intervals

 

[RickG]

Maybe you participants should examine the statistical terminology before patting yourselves on the back…

"accurate to approximately 3%"

that is covered here… "Accurate to" and "confidence levels" are distinct terminologies.

Statistics Glossary - confidence intervals

I admit, I haven't taken a statistics course since graduate school. But just one question Zaius. Do you ever read the sources you cite?
[FONT=helvetica, arial, ms sans serif]
[/FONT][FONT=helvetica, arial, ms sans serif]Confidence Level"[/FONT]
[FONT=helvetica, arial, ms sans serif]
[/FONT][FONT=helvetica, arial, ms sans serif]The confidence level is the probability value

associated with a confidence interval.



It is often expressed as a percentage. For example, say

, then the confidence level is equal to (1-0.05) = 0.95, i.e. a 95% confidence level.


Example

Suppose an opinion poll predicted that, if the election were held today, the Conservative party would win 60% of the vote. The pollster might attach a 95% confidence level to the interval 60% plus or minus 3%. That is, he thinks it very likely that the Conservative party would get between 57% and 63% of the total vote.
[/FONT]

 

[Zaius137]

Withdrawn

 

[verysincere]

Lev. 11:13, 19 And these are they which ye shall have in abomination among the fowls...And the stork, the heron after her kind, and the lapwing, and the bat. (See also Deut. 14:11, 18)

Bats aren't birds. They are mammals.

While I agree with most of your post and enjoyed how you admonished the ill-informed participants, you fell into several basic fallacies of your own in your traditional but flawed "bats aren't birds" criticism of the Biblical text.

1) Leviticus 11:13,19 never claims that "bats are birds". The Hebrew word involved can better be translated as "flying creatures"--- of which surely bats are among them---but that would make an awkward-sounding English translation, so "fowls" was an approximation decision made by the translators using the "King James English" of 1611. If you wish to contest the Hebrew exegesis of Bible translators who have been dead for centuries, go for it. But don't pretend that you've found "bats are birds" in the Biblical text. You didn't.

2) In any case, applying Linnaean taxonomy about a century before Carl Linnaeus was even born should have been obvious to you as AN ANACHRONISM FALLACY even if you are not well-informed about semantic fields and the basic linguistics of translating an ancient text.


As much as I enjoy the exposing of Young Earth Creationist pseudoscience and ignorance of the scriptures, it bothers me just as much when fundamentalists of another type make equally serious gaffes.

 

[Zaius137]

I admit, I haven't taken a statistics course since graduate school. But just one question Zaius. Do you ever read the sources you cite?
[FONT=helvetica, arial, ms sans serif]
[/FONT][FONT=helvetica, arial, ms sans serif]Confidence Level"[/FONT]
[FONT=helvetica, arial, ms sans serif]
[/FONT][FONT=helvetica, arial, ms sans serif]The confidence level is the probability value

associated with a confidence interval.



It is often expressed as a percentage. For example, say

, then the confidence level is equal to (1-0.05) = 0.95, i.e. a 95% confidence level.


Example

Suppose an opinion poll predicted that, if the election were held today, the Conservative party would win 60% of the vote. The pollster might attach a 95% confidence level to the interval 60% plus or minus 3%. That is, he thinks it very likely that the Conservative party would get between 57% and 63% of the total vote.
[/FONT]

What if all these electrical and gas layers are just results of gas diffusion and gravitational differentiation. When it is claimed in this paper that and accuracy is within 3% or 6% exactly what does that mean? A entire circular reasoning can be built around total speculation. It is simply an accuracy biased by assumptions that are highly speculative in the first place. Visual interpretation only agreed 5% of the time so you take a flawed assumption based on faulty interpretations saying that there is possibly some uniform gradient of ions or oxygen isotopes. That is simply un-provable and in fact pure speculation.

 

[RickG]

What if all these electrical and gas layers are just results of gas diffusion and gravitational differentiation.

electrical conductivity and acidity come from come from salts, not gases.

When it is claimed in this paper that and accuracy is within 3% or 6% exactly what does that mean? A entire circular reasoning can be built around total speculation. It is simply an accuracy biased by assumptions that are highly speculative in the first place.

In statistical analysis there are specific tests for testing the validity of the data and data sets. There is no circular reasoning, it's mathematics processing physical data, no assumptions.

Visual interpretation only agreed 5% of the time so you take a flawed assumption based on faulty interpretations saying that there is possibly some uniform gradient of ions or oxygen isotopes. That is simply un-provable and in fact pure speculation.

No, what the study said was:

Additional checks were also made, including having two stratigraphers from one group independently interpret selected core sections, and re-examination of of selected core sections. Overall, these tests indicated the manual interpretation of the visual observer was reproducible to about 5% in Holocene ice." (page 456)

​

Now, go back and read what your link said about confidence level. That 5% means a 95% reliability. Also, note that it was an independent observer. Two separate people processing the same data and only get a 5% difference in results. That is incredibly good.

 

[Zaius137]

Additional checks were also made, including having two stratigraphers from one group independently interpret selected core sections, and re-examination of of selected core sections. Overall, these tests indicated the manual interpretation of the visual observer was reproducible to about 5% in Holocene ice." (page 456)
Now, go back and read what your link said about confidence level. That 5% means a 95% reliability. Also, note that it was an independent observer. Two separate people processing the same data and only get a 5% difference in results. That is incredibly good.

“Overall, these tests indicated the manual interpretation of the visual observer was reproducible to about 5% in Holocene ice.”


This article was critical of visual interpretation of ice dates and states that agreements with that method only amounted to about 5%. It only shows how shaky the assumptions are with dating ice cores.

What is the oldest ice core on earth again?

8 million year old ice from Antarctica at Beacon Valley yields living micro organisms… HuH?


Fossil genes and microbes in the ol... [Proc Natl Acad Sci U S A. 2007] - PubMed - NCBI

 

[Zaius137]

A very young mountain range under the ice can not be explained by scientists… excuse me except by ad-hoc explanations… rift valley tectonics (just funny)

How Did Antarctica's Mysterious Ice-Encrusted Mountains Get There? - Gamburtsev Mountain Range - Popular Mechanics

 

[Zaius137]

8 million year old ice from Antarctica at Beacon Valley yields living micro organisms… Wait that is suspended animation (we can go to the stars now!) DNA with a half life of 1.1 million years…. They have done it! Or Antarctica has done it! Or maybe the ice is young.


Fossil genes and microbes in the ol... [Proc Natl Acad Sci U S A. 2007] - PubMed - NCBI

 

[Zaius137]

I admit, I haven't taken a statistics course since graduate school. But just one question Zaius. Do you ever read the sources you cite?
[FONT=helvetica, arial, ms sans serif]
[/FONT][FONT=helvetica, arial, ms sans serif]Confidence Level"[/FONT]
[FONT=helvetica, arial, ms sans serif]
[/FONT][FONT=helvetica, arial, ms sans serif]The confidence level is the probability value

associated with a confidence interval.



It is often expressed as a percentage. For example, say

, then the confidence level is equal to (1-0.05) = 0.95, i.e. a 95% confidence level.


Example

Suppose an opinion poll predicted that, if the election were held today, the Conservative party would win 60% of the vote. The pollster might attach a 95% confidence level to the interval 60% plus or minus 3%. That is, he thinks it very likely that the Conservative party would get between 57% and 63% of the total vote.
[/FONT]

In all seriousness Rick the alpha can not be your 3 or 6%...

About the alpha level…

Example:

“.01 alpha level means that you leave open a 1% possibility of being wrong if you reject the null hypothesis (which is kinda sorta simplifiable to a 99% chance that the alternate/research hypothesis is "correct"). If the alpha level is .05, there's a 5% chance of being wrong if you reject the null.”

Since the author uses “accurate to approximately 3%”

The alpha is not 3% but the result of (1-alpha) or an alpha of .97…

 

[RickG]

In all seriousness Rick the alpha can not be your 3 or 6%...

About the alpha level…

Example:

“.01 alpha level means that you leave open a 1% possibility of being wrong if you reject the null hypothesis (which is kinda sorta simplifiable to a 99% chance that the alternate/research hypothesis is "correct"). If the alpha level is .05, there's a 5% chance of being wrong if you reject the null.”

Since the author uses “accurate to approximately 3%”

The alpha is not 3% but the result of (1-alpha) or an alpha of .97…

Are you suggesting 0.97% rather than 97%?

 

[Zaius137]

Are you suggesting 0.97% rather than 97%?

Not at all…

It may be that the terminology used here does not carry strict statistical meaning.
But it is clear that alpha must be less than one.

Instead of saying “accurate to approximately 3%” it should say “inaccurate to approximately 3%” then you could say it would be the alpha in the strict sense.

 

[Lion Hearted Man]

In all seriousness Rick the alpha can not be your 3 or 6%...

About the alpha level…

Example:

“.01 alpha level means that you leave open a 1% possibility of being wrong if you reject the null hypothesis (which is kinda sorta simplifiable to a 99% chance that the alternate/research hypothesis is "correct"). If the alpha level is .05, there's a 5% chance of being wrong if you reject the null.”

Since the author uses “accurate to approximately 3%”

The alpha is not 3% but the result of (1-alpha) or an alpha of .97…

The alpha is 0.03, which very strongly favors rejection of the null hypothesis; it is often spoken of in percentage form...which would be a 3% chance of the null hypothesis being true if rejected.

An alpha of 0.97 would very strongly favor nonrejection of the null hypothesis.

 

[RickG]

Not at all…

It may be that the terminology used here does not carry strict statistical meaning.
But it is clear that alpha must be less than one.

Instead of saying “accurate to approximately 3%” it should say “inaccurate to approximately 3%” then you could say it would be the alpha in the strict sense.

It's not quite that simple, remember in the paper they used algorithms. Exactly what application they did not say, but knowing there were multiple variables it would have had to have been one dealing with least squares. Here's an example.

The standard algorithm for computing partial least squares regression components (i.e., factors) is nonlinear iterative partial least squares (NIPALS). There are many variants of the NIPALS algorithm which normalize or do not normalize certain vectors. The following algorithm, which assumes that the Xand Y variables have been transformed to have means of zero, is considered to be one of most efficient NIPALS algorithms.
For each h=1,…,c, where A0=X'Y, M0=X'X, C0=I, and c given,

1. compute qh, the dominant eigenvector of Ah'Ah
2. wh=ChAhqh, wh=wh/||wh||, and store wh into W as a column
3. ph=Mhwh, ch=wh'Mhwh, ph=ph/ch, and store ph into P as a column
4. qh=Ah'wh/ch, and store qh into Q as a column
5. Ah+1=Ah - chphqh' and Mh+1=Mh - chphph'
6. Ch+1=Ch - whph'

The factor scores matrix T is then computed as T=XW and the partial least squares regression coefficients B of Y on X are computed as B=WQ.

 

[Zaius137]

It's not quite that simple, remember in the paper they used algorithms. Exactly what application they did not say, but knowing there were multiple variables it would have had to have been one dealing with least squares. Here's an example.

The standard algorithm for computing partial least squares regression components (i.e., factors) is nonlinear iterative partial least squares (NIPALS). There are many variants of the NIPALS algorithm which normalize or do not normalize certain vectors. The following algorithm, which assumes that the Xand Y variables have been transformed to have means of zero, is considered to be one of most efficient NIPALS algorithms.
For each h=1,…,c, where A0=X'Y, M0=X'X, C0=I, and c given,

1. compute qh, the dominant eigenvector of Ah'Ah
2. wh=ChAhqh, wh=wh/||wh||, and store wh into W as a column
3. ph=Mhwh, ch=wh'Mhwh, ph=ph/ch, and store ph into P as a column
4. qh=Ah'wh/ch, and store qh into Q as a column
5. Ah+1=Ah - chphqh' and Mh+1=Mh - chphph'
6. Ch+1=Ch - whph'

The factor scores matrix T is then computed as T=XW and the partial least squares regression coefficients B of Y on X are computed as B=WQ.

This is the stuff I like… thanks.

Algorithms are fine but when they are not fettered by real parameters they hold no value. My experience as a programmer has given me a great appreciation for reality because an entire reality can be made to exist on only fantasy.

I would have liked to at least review the process in this paper.

 

[Jazer]

8 million year old ice from Antarctica at Beacon Valley yields living micro organisms…

Did you get the wrong link? Try this one: Eight-million-year-old bug is alive and growing
Eight-million-year-old bug is alive and growing - environment - 07 August 2007 - New Scientist

 

[cupid dave]

While I agree with most of your post and enjoyed how you admonished the ill-informed participants, you fell into several basic fallacies of your own in your traditional but flawed "bats aren't birds" criticism of the Biblical text.

1) Leviticus 11:13,19 never claims that "bats are birds". The Hebrew word involved can better be translated as "flying creatures"--- of which surely bats are among them---but that would make an awkward-sounding English translation, so "fowls" was an approximation decision made by the translators using the "King James English" of 1611. If you wish to contest the Hebrew exegesis of Bible translators who have been dead for centuries, go for it. But don't pretend that you've found "bats are birds" in the Biblical text. You didn't.

2) In any case, applying Linnaean taxonomy about a century before Carl Linnaeus was even born should have been obvious to you as AN ANACHRONISM FALLACY even if you are not well-informed about semantic fields and the basic linguistics of translating an ancient text.


As much as I enjoy the exposing of Young Earth Creationist pseudoscience and ignorance of the scriptures, it bothers me just as much when fundamentalists of another type make equally serious gaffes.

Good job.

People elsewhere have used the KJV translatiion to criticize the statement about whales and hares in much the same uneducated fashio.
I appreciate your research into scripture in the name if honest commmentary.

In regard to the flood, it ought at least be a reasonable cravate in the discussion to point out that the initial and predominate "audience" for scripture has been societies which could never have understood or entertained the idea that Modern man exploded out of Africa jsut 40,000 years ago, flooding up the mountains and filling every valley on earth.

Thr three racial stocks of a single Noah-type man to whom all people living today are related as the father in common 40 thusand years ago is hard to believe now.

That Genesis tells the story of the extinction of Neanderthal man and the whole strange truth we now have uncovered is not hard to see for those with an open mind and no bias either way.

 

[Zaius137]

Good job.

People elsewhere have used the KJV translatiion to criticize the statement about whales and hares in much the same uneducated fashio.
I appreciate your research into scripture in the name if honest commmentary.

In regard to the flood, it ought at least be a reasonable cravate in the discussion to point out that the initial and predominate "audience" for scripture has been societies which could never have understood or entertained the idea that Modern man exploded out of Africa jsut 40,000 years ago, flooding up the mountains and filling every valley on earth.

Thr three racial stocks of a single Noah-type man to whom all people living today are related as the father in common 40 thusand years ago is hard to believe now.

That Genesis tells the story of the extinction of Neanderthal man and the whole strange truth we now have uncovered is not hard to see for those with an open mind and no bias either way.

Well Dave I would like to mention a bit about the polymorphic aspects of the human genome that actually predicts a huge bottle neck of the human race happened. We creationists (Theists) knew this for several thousand years as Noah’s flood and just now science can actually see it in our genomes. That fairytale story of small populations over hundreds of thousands of years then mass extinction and then exponential population growth is not just nonsense but mathematically untenable. The out of Africa story 40,000 years ago is just that, a story.